Hands-On Docking With Molegro Virtual Docker

Abstract

Molegro Virtual Docker (MVD) integrates state-of-the-art search algorithms and scoring functions dedicated to protein-ligand docking simulations. It implements differential evolution as a search engine and MolDock and Plants scores to calculate binding affinity. In this work, we describe a workflow focused on how to build regression models to predict the inhibition of cyclin-dependent kinase 2 (CDK2). We employ available structural and binding data to construct machine learning models to calculate CDK2 inhibition based on the atomic coordinates obtained through docking simulations performed with MVD. We present a hands-on approach to show how to integrate docking results and machine learning methods available at Scikit-Learn to build targeted scoring functions. Our regression models show superior predictive performance compared with classical scoring functions. All CDK2 datasets and Jupyter Notebooks discussed in this work are available at GitHub: https://github.com/azevedolab/docking#readme. We made the source code of the program SAnDReS 2.0 available at https://github.com/azevedolab/sandres. © 2025 Elsevier B.V., All rights reserved.